Abstract
Purpose
Vitamin D has been implicated in lowering lung cancer risk, but serological data on the association among never-smoking women are limited. We report results examining the association of serum 25-hydroxyvitamin D [25(OH)D] concentrations with lung cancer risk among female never smokers. We also examined whether the association was modified by vitamin D supplementation and serum vitamin A concentrations.
Methods
In the Women’s Health Initiative, including the Calcium/Vitamin D (CaD) Trial, we selected 298 incident cases (191 non-small-cell lung cancer [NSCLC] including 170 adenocarcinoma) and 298 matched controls of never smokers. Baseline serum 25(OH)D was assayed by a chemiluminescent method. Logistic regression was used to estimate odds ratios (ORs) for quartiles and pre-defined clinical cutoffs of serum 25(OH)D concentrations.
Results
Comparing quartiles 4 versus 1 of serum 25(OH)D concentrations, ORs were 1.06 (95% confidence interval [CI]=0.61-1.84) for all lung cancer, 0.94 (95% CI=0.52-1.69) for NSCLC, and 0.91 (95% CI=0.49-1.68) for adenocarcinoma. Comparing serum 25(OH)D ≥75 (high) versus <30 nmol/L (deficient), ORs were 0.76 (95% CI=0.31-1.84) for all lung cancer, 0.71 (95% CI=0.27-1.86) for NSCLC, and 0.81 (95% CI=0.31-2.14) for adenocarcinoma. There is suggestive evidence that CaD supplementation (1g calcium+400 IU D3/day) and a high level of circulating vitamin A may modify associations of 25(OH)D with lung cancer overall and subtypes (P-interaction <0.10).
Conclusions
In this group of never-smoking postmenopausal women, the results did not support the hypothesis of an association between serum 25(OH)D and lung cancer risk.
Keywords: 25-hydroxyvitamin D, lung cancer, postmenopausal women, never smokers, histology
INTRODUCTION
Carcinogens from environmental exposures, including second-hand tobacco smoke, are closely tied to lung cancers occurring in never smokers, but a large fraction of these lung cancers may be attributable to other unknown risk factors [1]. Low vitamin D status has been hypothesized as a risk factor for lung cancer because there are abundant vitamin D metabolic enzymes as well as the vitamin D receptor (VDR) expressed in the lung [2, 3]. Humans obtain vitamin D from diet and through cutaneous synthesis as ultraviolet B radiation exposure converts 7-dehydrocholesterol to vitamin D. Through enzymatic reactions in the liver and kidney, vitamin D is converted to 25-hydroxyvitamin D [25(OH)D], the standard biomarker for assessing vitamin D status, and further to 1,25-dihydroxyvitamin D [4]. 1,25-dihydroxyvitamin D has anti-proliferative properties [5]. 1,25-dihydroxyvitamin D also binds to the VDR, pairing with retinoid X receptor (RXR) to regulate transcription of target genes that control proliferation, apoptosis, and angiogenesis [6, 7].
Although epidemiological data are inconsistent on the association of 25(OH)D concentrations with lung cancer risk (Supplementary Table 1) [8–17], a meta-analysis of 10 prospective studies showed an inverse, dose-response relationship (5% reduction in incidence or mortality for each 10 nmol/L increment) [18]. To date, two studies have reported findings for women [8, 9], of which only one study examined the association of 25(OH)D concentrations with lung cancer incidence [8]. Serum 25(OH)D concentrations ≥47 versus ≤30 nmol/L were associated with a lower lung cancer risk in a cohort study of Finnish women (n=3,730) (relative risk=0.16, 95% confidence interval [CI]=0.04-0.59) [8]; however, the number of lung cancer cases was very small (n=25). Data from the Third National Health and Nutrition Examination Survey (NHANES III) showed no association of serum 25(OH)D with lung cancer mortality in women (hazard ratio [HR]=0.64, 95% CI=0.23-1.79, ≥80 vs. <50 nmol/L) [9]. However, in another analysis of the NHANES III data, a significant decrease in lung cancer mortality with higher serum 25(OH)D was observed in never smokers and former smokers who quit for 20 years or longer and had higher serum 25(OH)D levels (HR=0.31, 95% CI=0.13-0.76, >80 versus <44 nmol/L) [10]. Lung cancer incidence was not assessed in the NHANES III and, in some participants, blood samples might have been obtained after the lung cancer diagnosis. Despite these limitations, the findings suggest that serum 25(OH)D may also play a role in lung cancer etiology in never smokers. Research is needed on the association of 25(OH)D concentrations with lung cancer risk among female never smokers for whom the risk factors remain elusive.
Here, we report results examining the association of serum 25(OH)D concentrations with lung cancer risk among never smokers in the Women’s Health Initiative (WHI). WHI is a large study of 161,808 postmenopausal women in the U.S. and includes an Observational Study (OS) and three overlapping clinical trials (CTs) [19]. Among CTs, the Calcium/Vitamin D (CaD) Trial was a randomized, placebo-controlled trial testing supplementation with daily 1g of calcium carbonate and 400 IU of vitamin D3 for preventing fracture and colorectal cancer [20, 21]. Thus, we also examined whether the association of serum 25(OH)D concentrations with lung cancer risk differed by the CaD Trial intervention. In addition, we investigated whether serum vitamin A (retinol), which supplies ligands necessary for the VDR-RXR heterodimer, modified the relationship between serum 25(OH)D and lung cancer risk.
METHODS
Study population
We conducted a case-control study nested in the WHI CTs and OS [19, 22]. The eligibility criteria for both the CTs and the OS included being postmenopausal, aged 50–70 years with a life expectancy of at least three years, no history of breast cancer, and no history of other cancers within the previous ten years. At baseline (1993–98), women were enrolled in the Hormonal Therapy Trials (n=27,347) and Dietary Modification Trial (n=48,835). After one year, CT participants were invited to join the CaD Trial (n=36,282). For the CaD Trial, women who had a history of hypercalcemia, kidney stones, or corticosteroid or calcitriol use and women who chose to continue vitamin D supplementation over 600 IU/d were ineligible. Personal use of calcium and vitamin D supplements (initially up to 600 IU/d, and later, 1000 IU/d) was permitted during the trial. Women either ineligible or unwilling to participate in the CTs were enrolled in the OS for prospective follow-up (n=93,676). The protocol was approved by the institutional review boards of each clinical center and the WHI Clinical Coordinating Center. All women provided written informed consent. The current study was approved by the institutional review board of the Fred Hutchinson Cancer Research Center.
Outcome assessment and case and control selection
Participants reported all new cancer diagnoses, including lung cancer, semiannually in the CTs and annually in the OS until 2005 when the outcomes were thereafter reported annually. Trained study physicians, blinded to WHI study components, at local clinics confirmed and adjudicated cases by reviewing medical records [23]. Tumor histology and stage were coded using the International Classification of Disease for Oncology, second edition. For the current study, inclusion criteria were cases who were never smokers and had adequate serum samples at baseline. Smoking status was based on self-report assessed by a standardized questionnaire [24], and never smokers were defined as those who had not smoked >100 cigarettes in their lifetime and did not smoke at the time of questionnaire completion [25]. A total of 369 incident lung cancer cases out of 2,491 ascertained as of September 30, 2012 were eligible; from these cases, 300 were randomly selected. Controls were women who never smoked, provided baseline serum samples, and did not have a diagnosis of cancer other than non-melanoma skin cancer up to the time of the case’s diagnosis date. One control was selected for each case from the risk set at the time of the case’s event, with additional matching on age (±1 year) and study (CT/OS). Two case-control pairs were excluded due to missing data on serum 25(OH)D concentrations and body mass index (BMI), leaving 298 case-control pairs for statistical analyses.
Blood sample processing and analysis
Whole blood was collected after an overnight, twelve-hour fast at baseline. The samples were processed to serum, aliquoted in dim/yellow light at the clinical centers and stored at –70°C within two hours of collection and at –80°C for long-term storage. Serum 25(OH)D was assayed in duplicates using LIAISON® 25 OH Vitamin D TOTAL Assay (DiaSorin, Stillwater, MN), a chemiluminescent assay providing the combined concentrations of both 25(OH)D2 and 25(OH)D3. Serum retinol and retinyl palmitate, the main component of retinyl esters [26, 27], were measured by high-performance liquid chromatography as described previously [28]. We did not assay other retinyl esters because their serum concentrations were very low in WHI participants. All assays were performed at the Fred Hutchinson Cancer Research Center Biomarker Laboratory, which participates in the National Institute of Standards and Technology Micronutrients Measurement Quality Assurance Program for fat-soluble vitamins and carotenoids in human serum and the Vitamin D Metabolites Quality Assurance Program. Cases and matched controls were analyzed in the same batch; laboratory staff was blinded to the case/control status of samples. Average inter-assay coefficients of variation (CV) were 5.5% for 25(OH)D, 2.6% for retinol, and 14.6% for retinyl palmitate in 30 pairs of blinded duplicate samples.
Questionnaire and anthropometric data collection
Demographic characteristics, including age, race/ethnicity, education, history of chronic diseases, and health-related behaviors, including physical activity, smoking history, and use of postmenopausal hormones and other medications, were collected at baseline by standardized questionnaires [19]. OS participants additionally reported 1) whether they ever or currently lived and worked with a smoker after the age of 18 years, and 2) numbers of years living and working with a smoker, as an assessment of second-hand smoke exposure. Recreational physical activity was estimated as metabolic equivalent task (MET)-hours per week. Dietary intakes were assessed at baseline by a self-administered, 122-item food frequency questionnaire, validated by four 24-hour dietary recalls and a 4-day food record [29]. Supplemental vitamin use was assessed by a simplified inventory method [30]. Weight and height were measured at baseline clinic visits by trained staff using a standardized protocol; BMI was computed as [weight (kg)/height2 (m2)].
Statistical analysis
Differences in baseline levels of exposure variables and potential confounders between cases and controls were examined with t-test (continuous variables) and chi-square test (categorical variables). Correlations were evaluated between serum concentrations of 25(OH)D, retinol, and retinyl palmitate and intake values of vitamin D and vitamin A from diet and supplements. We used conditional logistic regression models to estimate odds ratios (ORs) and 95% CIs for lung cancer risk. Cases and controls were grouped by quartiles of serum 25(OH)D concentrations; the quartile cutoffs were selected based on the distribution in control participants. In addition, we used season-specific quartiles to examine the serum 25(OH)D-lung cancer association as a secondary analysis [31]. We also tested the 25(OH)D-lung cancer association using a priori selected clinical cut points serum 25(OH)D concentrations defined as deficient (<30 nmol/L), inadequate (30 to <50 nmol/L), sufficient (50 to <75 nmol/L), and high (≥75 nmol/L). The cutoffs for deficient, inadequate, and sufficient statuses were based on the Institute of Medicine classification [32], and the high status has been suggested to be beneficial for various health outcomes including cancer [33]. Tests of linear trend across increasing categories of serum 25(OH)D concentrations were conducted using the Wald tests by modeling the median values of each category as a single continuous variable in the models. The selection of covariates was based on biomedical knowledge and their influence in risk estimates [1, 34]. Covariates in the final regression model included race/ethnicity (black/African American, Hispanic/Latino, non-Hispanic white, others), baseline body mass index (<25, 25 to <30, 30 to <35, ≥35 kg/m2), the CaD Trial randomization assignment (yes, no), serum retinol concentrations (quartiles), and season of blood draw (summer, winter). Other potential risk factors for lung cancer including recreational physical activity, use of postmenopausal hormonal replacement therapy, Hormonal Therapy Trials randomization assignments, and history of lung or respiratory tract diseases were not included in the final model because they did not change risk estimates. Unconditional logistic regression was used in the analyses for histological subtypes, i.e., NSCLC and adenocarcinoma, to maximize the sample size. In sensitivity analyses, we evaluated whether preclinical cancer status affects serum 25(OH)D concentrations by excluding lung cancer cases diagnosed within two years after the baseline blood draw and their matched controls. Because African-American women have lower serum 25(OH)D concentrations than women of other races in general [35], we also examined the association of serum 25(OH)D with lung cancer risk after excluding African Americans. Additionally, we adjusted second-hand smoke exposure for OS participants in whom this information was collected.
To evaluate the hypothesized effect modification, we stratified the association of serum 25(OH)D concentrations by (a) the CaD Trial intervention and (b) circulating vitamin A, i.e., serum concentrations of retinol and retinyl palmitate and the ratio of retinyl palmitate to retinol because the serum retinyl palmitate/retinol ratio can indicate excess circulating vitamin A [27]. For the analysis of the CaD Trial, the intervention group was women who were randomized to the intervention arm of the trial. The no intervention group included women who were randomized to the placebo arm of the trial and those who did not participate in the trial; the average intake values of vitamin D from supplements at baseline were similar between these two groups (186 IU/d and 174 IU/d, respectively). We grouped participants as below or above the sufficient level (50 nmol/L) of serum 25(OH)D, instead of using a finer cutoff, because of the relatively small number of lung cancer cases that occurred in the intervention arm (n=31). For evaluating the effect modification of serum vitamin A, we grouped participants using the quartiles of serum 25(OH)D concentrations and stratified by the median values of serum concentrations of retinol and retinyl palmitate and the retinyl palmitate/retinol ratio to maintain an approximately even number of participates between groups. The interaction was assessed using Wald tests (1 df) of the cross-product term between serum 25(OH)D categories and CaD Trial intervention or above/below the median of serum retinol and retinyl palmitate concentrations and the retinyl palmitate/retinol ratio. Statistical analyses were conducted using SAS 9.4 (Cary, NC).
RESULTS
Among cases, the majority of histological subtype was non-small cell lung cancer (NSCLC) (n=191; 72% among cases with known histology); within NSCLC, adenocarcinoma was the most common subtype (n=170; 89% of NSCLC) (Supplementary Table 2). Cases and controls did not differ with regard to demographic and lifestyle characteristics, history of lung diseases and hormone use, or serum biomarkers (Table 1). Serum 25(OH)D concentrations were higher among women who had the blood draw in summer than among those who had the blood draw in winter but there was no difference between the cases and controls. Serum 25(OH)D concentrations had a modest correlation with total vitamin D intake from diet and supplements (r=0.34; Supplementary Table 3).
Table 1.
Baseline characteristics of lung cancer cases and controls in the WHI CTs and OS
| Characteristic | Cases (n=298) | Controls (n=298) | P-value1 |
|---|---|---|---|
| Age, years (mean ± SD) | 65.5 ± 7.1 | 65.6 ± 7.1 | 1.00 |
| Region (%) | |||
| Northeast | 72 (24.2) | 67 (22.5) | 0.70 |
| South | 72 (24.2) | 77 (25.8) | |
| Midwest | 62 (20.8) | 71 (23.8) | |
| West | 92 (30.9) | 83 (27.9) | |
| Race-ethnicity (%) | 0.63 | ||
| Black/African American | 18 (6.0) | 17 (5.7) | |
| Hispanic/Latino | 7 (2.3) | 10 (3.3) | |
| Non-Hispanic white | 254 (85.3) | 258 (86.6) | |
| Others | 19 (6.4) | 13 (4.4) | |
| Education (%) | 0.63 | ||
| High school or less | 12 (4.0) | 17 (5.7) | |
| School after high school | 153 (51.3) | 159 (53.4) | |
| College degree or higher | 131 (44.0) | 119 (39.9) | |
| Unknown | 2 (0.7) | 3 (1.0) | |
| Body-mass index, kg/m2 (mean ± SD) | 27.1 ± 5.4 | 27.4 ± 5.4 | 0.61 |
| Recreational physical activity2, MET-h/week (mean ± SD) | 13.4 ± 14.4 | 13.0 ± 12.4 | 0.74 |
| Ever living with a smoker in adulthood3 (%) | 103 (58.5) | 109 (61.9) | 0.51 |
| Ever working with a smoker in adulthood3 (%) | 123 (69.9) | 113 (64.2) | 0.26 |
| Alcohol use (%) | 0.67 | ||
| Non drinkers | 50 (16.8) | 56 (18.8) | |
| Past drinkers | 44 (14.8) | 51 (17.1) | |
| Current drinkers | 202 (67.8) | 190 (63.8) | |
| Oral contraceptive use (ever, %) | 103 (34.6) | 100 (33.6) | 0.80 |
| Use of hormone therapy (%) | 0.44 | ||
| Never | 125 (41.9) | 139 (46.6) | |
| Past | 55 (18.5) | 43 (14.4) | |
| Current, estrogen alone | 71 (23.8) | 75 (25.2) | |
| Current estrogen + progesterone | 47 (15.8) | 41 (13.8) | |
| Participated in Hormone Therapy Trial (%) | 55 (18.5) | 51 (17.1) | 0.67 |
| Participated in Calcium/vitamin D Trial (%) | 58 (19.5) | 69 (23.2) | 0.27 |
| History of lung disease4 (ever, %) | 34 (11.4) | 24 (8.1) | 0.17 |
| Serum 25(OH)D, nmol/L (mean ± SD) [Median (IQR)] |
47.6 ± 16.2 47.4 (35.5-59.3) |
47.1 ± 17.8 45.3 (33.9-58.9) |
0.68 |
| Blood draw in summer5 (mean ± SD) [Median (IQR)] |
50.4 ± 16.3 50.8 (37.9-63.3) |
48.0 ± 17.8 46.6 (34.6-59.6) |
0.22 |
| Blood draw in winter5 (mean ± SD) [Median (IQR)] |
44.8 ± 15.8 43.1 (33.4-56.5) |
45.9 ± 17.8 43.9 (32.6-58.3) |
0.57 |
| Serum retinol, μg/dL (mean ± SD) [Median (IQR)] |
67.6 ± 14.2 65.8 (58.1-76.2) |
66.4 ± 15.4 64.4 (55.8-75.7) |
0.37 |
| Serum retinyl palmitate, μg/dL (mean ± SD) [Median (IQR)] |
4.08 ± 3.05 3.47 (2.48-4.65) |
3.92 ± 3.74 2.99 (2.29-4.30) |
0.54 |
| Serum retinyl palmitate/retinol ratio, % (mean ± SD) [Median (IQR)] |
6.1 ± 4.6 5.1 (2.8-7.2) |
5.9 ± 5.6 4.7 (3.7-6.2) |
0.67 |
IQR, inter-quartile range; SD, standard deviation
t-tests for continuous variables and chi-square tests for categorical variables
9 cases and 11 controls had missing values.
Assessed only among women in the OS (176 case-control pairs)
Ever-diagnosed for asthma and/or emphysema
Summer months were May, June, July, August, September, and October; winter months were November, December, January, February, March, and April.
The unadjusted model showed that there was no association between serum 25(OH)D in quartiles and lung cancer, NSCLC, and adenocarcinoma risk (Table 2). After adjusting for the covariates, there remained no association between serum 25(OH)D and any lung cancer. Women in quartile (Q) 4 had adjusted ORs of 1.06 (95% CI=0.61-1.84) for all lung cancer, 0.94 (95% CI=0.52-1.69) for NSCLC, and 0.91 (95% CI=0.49-1.68) for adenocarcinoma, compared with women in Q1. In the secondary analysis, similar risk estimates were derived from season-specific quartiles: adjusted OR of Q4 versus Q1 = 1.06 (95% CI=0.60-1.85) for all lung cancer, 1.02 (95% CI=0.56-1.84) for NSCLC, and 1.00 (95% CI=0.54-1.85) for adenocarcinoma (data not shown). Also, there was no association after we excluded cases within two years after blood draw and their matched controls (adjusted OR=0.97, 95% CI=0.51-1.84, Q4 vs. Q1) or African Americans (adjusted OR=1.15, 95% CI=0.68-1.94, Q4 vs. Q1) (data not shown). In the OS, additional adjustment of ever living and working with a smoker in adulthood did not change the risk estimates for lung cancer (without the adjustment: OR=1.11, 95% CI=0.54-2.29; with the adjustment: 1.16, 95% CI=0.55-2.46; both Q4 vs. Q1; data not shown). For the models with clinical cutoffs (Table 3), no association with lung cancer was observed for women with ≥75 nmol/L compared with those with <30 nmol/L of serum 25(OH)D concentrations (adjusted OR=0.76, 95% CI=0.31-1.84). Similar ORs were also observed for NSCLC (adjusted OR=0.71, 95% CI=0.27-1.86) and adenocarcinoma (adjusted OR=0.81, 95% CI=0.31-2.14) among women with high vs. deficient levels of serum 25(OH)D.
Table 2.
Odds ratios for lung cancer risk according to the quartiles of serum 25(OH)D concentrations
| Serum 25(OH)D | N (cases) |
N (controls) |
Unadjusted OR (95% CI) |
Adjusted OR (95% CI) |
P-trend | |
|---|---|---|---|---|---|---|
| All lung cancer1 | ||||||
| Quartile 14 (Ref.) | 64 | 73 | 1.00 | 1.00 | 0.85 | |
| Quartile 2 | 74 | 75 | 1.13 (0.40-1.81) | 1.11 (0.65-1.88) | ||
| Quartile 3 | 84 | 75 | 1.27 (0.81-2.00) | 1.15 (0.69-1.90) | ||
| Quartile 4 | 76 | 75 | 1.16 (0.72-1.87) | 1.06 (0.61-1.84) | ||
|
| ||||||
| Non-small-cell lung cancer2,3 | ||||||
| Quartile 1 (Ref.) | 43 | 73 | 1.00 | 1.00 | 0.78 | |
| Quartile 2 | 49 | 75 | 1.11 (0.66-1.87) | 1.07 (0.61-1.88) | ||
| Quartile 3 | 54 | 75 | 1.22 (0.73-2.04) | 1.05 (0.60-1.85) | ||
| Quartile 4 | 45 | 75 | 1.02 (0.60-1.73) | 0.94 (0.52-1.69) | ||
|
| ||||||
| Adenocarcinoma2 | ||||||
| Quartile 1 (Ref.) | 39 | 73 | 1.00 | 1.00 | 0.81 | |
| Quartile 2 | 42 | 75 | 1.05 (0.61-1.80) | 0.98 (0.55-1.76) | ||
| Quartile 3 | 50 | 75 | 1.25 (0.74-2.12) | 1.05 (0.58-1.88) | ||
| Quartile 4 | 39 | 75 | 0.97 (0.56-1.68) | 0.91 (0.49-1.68) | ||
N, number of participants; OR, odds ratio; CI, confidence interval
Odds ratios were estimated by conditional logistic regression models, adjusted for race/ethnicity, BMI, CaD Trial allocations, serum retinol concentrations, and season of blood draw.
Odds ratios were estimated by unconditional logistic regression, adjusted for the covariates and matching factors (age and CT/OS status).
Non-small cell lung cancer included adenocarcinoma, squamous cell carcinoma, and large cell carcinoma.
Serum 25(OH)D concentrations cutoffs were Q1: <33.9, Q2: 33.9-<45.3, Q3: 45.3-<58.9, Q4: ≥58.9 nmol/L.
Table 3.
Associations of serum 25(OH)D concentrations, categorized as pre-defined clinical cutoffs, with lung cancer risk
| Serum 25(OH)D, nmol/L | N (cases) |
N (controls) |
Unadjusted OR (95% CI) |
Adjusted OR (95% CI) |
P-trend | |
|---|---|---|---|---|---|---|
| All lung cancer1 | ||||||
| <30 (Ref.) | 43 | 55 | 1.00 | 1.00 | 0.91 | |
| 30- <50 | 120 | 121 | 1.31 (0.82-2.09) | 1.22 (0.72-2.05) | ||
| 50- <75 | 121 | 102 | 1.57 (0.96-2.57) | 1.34 (0.77-2.35) | ||
| ≥75 | 14 | 20 | 0.86 (0.38-1.90) | 0.76 (0.31-1.84) | ||
|
| ||||||
| Non-small-cell lung cancer2,3 | ||||||
| <30 (Ref.) | 31 | 55 | 1.00 | 1.00 | 0.69 | |
| 30- <50 | 79 | 121 | 1.16 (0.69-1.96) | 0.99 (0.56-1.76) | ||
| 50- <75 | 72 | 102 | 1.25 (0.73-2.14) | 1.01 (0.55-1.87) | ||
| ≥75 | 9 | 20 | 0.80 (0.32-1.97) | 0.71 (0.27-1.86) | ||
|
| ||||||
| Adenocarcinoma2 | ||||||
| <30 (Ref.) | 28 | 55 | 1.00 | 1.00 | 0.88 | |
| 30- <50 | 69 | 121 | 1.12 (0.65-1.93) | 0.92 (0.51-1.67) | ||
| 50- <75 | 64 | 102 | 1.23 (0.71-2.14) | 0.99 (0.53-1.86) | ||
| ≥75 | 9 | 20 | 0.88 (0.36-2.19) | 0.81 (0.31-2.14) | ||
N, number of participants; OR, odds ratio; CI, confidence interval
Odds ratios were estimated by conditional logistic regression models, adjusted for race/ethnicity, BMI, CaD Trial allocations, and serum retinol concentrations, and season of blood draw.
Odds ratios were estimated by unconditional logistic regression, adjusted for the covariates and matching factors (age and CT/OS status).
Non-small cell lung cancer included adenocarcinoma, squamous cell carcinoma, and large cell carcinoma.
Among women with serum 25(OH)D concentrations <50 nmol/L, the CaD Trial intervention versus no intervention was not associated with lung cancer risk (OR=0.99, 95% CI=0.52-1.90, Table 4). Women with serum 25(OH)D ≥50 nmol/L at baseline and receiving the CaD intervention there was a trend for lower lung cancer risk (OR=0.42, 95% CI=0.16-1.14), compared with those with serum 25(OH)D <50 nmol/L at baseline and no exposure to the intervention supplements (P-interaction=0.07). The ORs were 0.39 (95% CI=0.12-1.29; P-interaction=0.09) for NSCLC and 0.35 (95% CI=0.09-1.30; P-interaction=0.06) for adenocarcinoma.
Table 4.
Effect modification of the Calcium/Vitamin D (CaD) Trial intervention on the association of serum 25(OH)D concentrations with lung cancer risk
| Calcium/vitamin D Trial3 | ||||
|---|---|---|---|---|
|
| ||||
| Serum 25(OH)D | No intervention | Intervention | P-interaction | |
| All lung cancer1 | ||||
| <50 nmol/L | N (cases/controls) | 139/152 | 24/24 | |
| OR (95% CI) | 1.00 (Referent) | 0.99 (0.52-1.90) | 0.07 | |
| ≥50 nmol/L | N (cases/controls) | 128/107 | 7/15 | |
| OR (95% CI) | 1.22 (0.84-1.77) | 0.42 (0.16-1.14) | ||
|
| ||||
| Non-small cell lung cancer2 | ||||
| <50 nmol/L | N (cases/controls) | 93/152 | 17/24 | |
| OR (95% CI) | 1.00 (Referent) | 1.14 (0.54-2.39) | 0.09 | |
| ≥50 nmol/L | N (cases/controls) | 77/107 | 4/15 | |
| OR (95% CI) | 1.13 (0.74-1.72) | 0.39 (0.12-1.29) | ||
|
| ||||
| Adenocarcinoma2 | ||||
| <50 nmol/L | N (cases/controls) | 82/152 | 15/24 | |
| OR (95% CI) | 1.00 (Referent) | 1.20 (0.55-2.60) | 0.06 | |
| ≥50 nmol/L | N (cases/controls) | 70/107 | 3/15 | |
| OR (95% CI) | 1.20 (0.77-1.86) | 0.35 (0.09-1.30) | ||
Odds ratios were estimated by conditional logistic regression models, adjusted for race/ethnicity, BMI, serum retinol concentrations, and season of blood draw.
Odds ratios were estimated by unconditional logistic regression, adjusted for the covariates and matching factors (age and CT/OS status).
The no intervention group included women who were randomly assigned to the placebo arm in the CaD trial and those who did not participate in the trial. The intervention group included women who were randomly assigned to the calcium and vitamin D arm in the CaD trial.
There was no clear pattern of effect modification of serum retinol, retinyl palmitate, and the retinyl palmitate/retinol ratio on the association of serum 25(OH)D with the overall risk of lung cancer (Table 5). Among women with lower serum retinol palmitate/retinol ratio (below median), those with higher versus lower serum 25(OH)D tended to have a lower risk of NSCLC (OR=0.62, 95% CI=0.28-1.39 in Q4 versus Q1 of 25(OH)D). The pattern was similar for adenocarcinoma. However, women with both higher serum 25(OH)D and high retinyl palmitate/retinol ratio tended to have a higher risk of lung cancer, compared with women with both low serum 25(OH)D concentrations (quartile 1) and low retinyl palmitate/retinol ratio (p-interaction=0.06 for both NSCLC and adenocarcinoma).
Table 5.
Effect modification of serum retinol, retinyl palmitate, and the retinyl palmitate/retinol ratio on the association of serum 25(OH)D concentrations with lung cancer risk
| Serum 25(OH)D concentrations
|
P-interaction | |||||
|---|---|---|---|---|---|---|
| Q1 | Q2 | Q3 | Q4 | |||
| All lung cancer1 | ||||||
| Serum retinol | ||||||
| Below median (<66 μg/dL) | N, case/control | 39/44 | 30/37 | 34/27 | 27/34 | 0.80 |
| OR (95% CI) | 1.00 (Referent) | 0.92 (0.48-1.78) | 1.44 (0.72-2.86) | 0.98 (0.47-2.04) | ||
| Above median (≥66 μg/dL) | N, case/control | 25/29 | 44/38 | 50/48 | 49/41 | |
| OR (95% CI) | 0.97 (0.46-2.02) | 1.41 (0.70-2.82) | 1.19 (0.65-2.20) | 1.41 (0.72-2.76) | ||
| Serum retinyl palmitate | ||||||
| Below median (<3.0 μg/dL) | N, case/control | 37/51 | 33/40 | 26/28 | 23/32 | 0.80 |
| OR (95% CI) | 1.00 (Referent) | 1.11 (0.57-2.16) | 1.25 (0.62-2.54) | 1.03 (0.51-2.06) | ||
| Above median (≥3.0 μg/dL) | N, case/control | 27/22 | 41/35 | 58/47 | 53/43 | |
| OR (95% CI) | 1.77 (0.83-3.75) | 1.74 (0.91-3.36) | 1.74 (0.96-3.17) | 1.75 (0.92-3.35) | ||
| Serum retinyl palmitate/retinol ratio | ||||||
| Below median (<0.05) | N, case/control | 39/47 | 38/41 | 37/39 | 31/37 | 0.81 |
| OR (95% CI) | 1.00 (Referent) | 1.18 (0.61-2.29) | 1.17 (0.61-2.24) | 1.14 (0.59-2.21) | ||
| Above median (≥0.05) | N, case/control | 25/26 | 36/34 | 47/36 | 45/38 | |
| OR (95% CI) | 1.22 (0.61-2.46) | 1.35 (0.70-2.60) | 1.64 (0.88-3.08) | 1.51 (0.76-3.00) | ||
|
| ||||||
| Non-small cell lung cancer2 | ||||||
| Serum retinol | ||||||
| Below median (<66 μg/dL) | N, case/control | 26/44 | 18/37 | 21/27 | 13/34 | 0.43 |
| OR (95% CI) | 1.00 (Referent) | 0.84 (0.39-1.83) | 1.23 (0.56-2.68) | 0.68 (0.29-1.58) | ||
| Above median (≥66 μg/dL) | N, case/control | 17/29 | 31/38 | 33/48 | 32/41 | |
| OR (95% CI) | 0.95 (0.43-2.11) | 1.37 (0.67-2.81) | 1.16 (0.57-2.34) | 1.37 (0.66-2.82) | ||
| Serum retinyl palmitate | ||||||
| Below median (<3.0 μg/dL) | N, case/control | 28/51 | 21/40 | 15/28 | 11/32 | 0.30 |
| OR (95% CI) | 1.00 (Referent) | 0.94 (0.45-1.98) | 0.95 (0.42-2.15) | 0.66 (0.28-1.58) | ||
| Above median (≥3.0 μg/dL) | N, case/control | 15/22 | 28/35 | 39/47 | 34/43 | |
| OR (95% CI) | 1.16 (0.50-2.71) | 1.49 (0.73-3.05) | 1.47 (0.75-2.91) | 1.52 (0.76-3.07) | ||
| Serum retinyl palmitate/retinol ratio | ||||||
| Below median (<0.05) | N, case/control | 30/47 | 26/41 | 23/39 | 14/37 | 0.06 |
| OR (95% CI) | 1.00 (Referent) | 0.94 (0.46-1.91) | 0.86 (0.41-1.78) | 0.62 (0.28-1.39) | ||
| Above median (≥0.05) | N, case/control | 13/26 | 23/34 | 31/36 | 31/38 | |
| OR (95% CI) | 0.69 (0.30-1.59) | 1.07 (0.51-2.22) | 1.29 (0.64-2.63) | 1.30 (0.64-2.65) | ||
|
| ||||||
| Adenocarcinoma2 | ||||||
| Serum retinol | ||||||
| Below median (<66 μg/dL) | N, case/control | 23/44 | 16/37 | 19/27 | 11/34 | |
| OR (95% CI) | 1.00 (Referent) | 0.84 (0.36-1.82) | 1.26 (0.56-2.82) | 0.66 (0.27-1.61) | 0.44 | |
| Above median (≥66 μg/dL) | N, case/control | 16/29 | 27/38 | 31/48 | 28/41 | |
| OR (95% CI) | 1.03 (0.46-2.32) | 1.35 (0.64-2.08) | 1.25 (0.61-2.57) | 1.42 (0.67-3.01) | ||
| Serum retinyl palmitate | ||||||
| Below median (<3.0 μg/dL) | N, case/control | 25/51 | 16/40 | 12/28 | 11/32 | 0.40 |
| OR (95% CI) | 1.00 (Referent) | 0.77 (0.35-1.70) | 0.80 (0.33-1.91) | 0.73 (0.30-1.77) | ||
| Above median (≥3.0 μg/dL) | N, case/control | 14/22 | 26/35 | 38/47 | 28/43 | |
| OR (95% CI) | 1.13 (0.47-2.70) | 1.53 (0.74-3.19) | 1.61 (0.80-3.22) | 1.44 (0.69-2.99) | ||
| Serum retinyl palmitate/retinol ratio | ||||||
| Below median (<0.05) | N, case/control | 27/47 | 22/41 | 20/39 | 13/37 | 0.06 |
| OR (95% CI) | 1.00 (Referent) | 0.85 (0.41-1.77) | 0.79 (0.37-1.69) | 0.63 (0.28-1.46) | ||
| Above median (≥0.05) | N, case/control | 12/26 | 20/34 | 30/36 | 26/38 | |
| OR (95% CI) | 0.67 (0.28-1.58) | 1.03 (0.48-2.20) | 1.41 (0.68-2.90) | 1.25 (0.60-2.62) | ||
Odds ratios were estimated by conditional logistic regression, adjusted for matching factors (age and study status), race/ethnicity, BMI, CaD Trial allocation, and season of blood draw.
Odds ratios were estimated by unconditional logistic regression, adjusted for the covariates and matching factors (age and CT/OS status).
DISCUSSION
In this sample of never-smoking postmenopausal women, the data did not support the hypothesis that higher versus lower serum 25(OH)D concentrations were associated with a lower risk of lung cancer. Our effect modification analysis for the CaD Trial intervention showed a trend towards a lower yet non-significant risk of lung cancer was observed for women with a higher serum 25(OH)D concentrations and receiving vitamin D supplementation, compared with those with lower serum 25(OH)D concentrations and not receiving the trial supplement. It is important to note that there were very few cases and controls with higher serum 25(OH)D concentrations at baseline and in the intervention arm so the finding must be interpreted with caution. The CaD supplementation increased serum 25(OH)D concentration on average from 45.5 nmol/L in the placebo arm to 60.8 nmol/L in the active intervention arm at two years of post-randomization [36]. However, the vitamin D supplement dose used in the trial (400 IU/d) might not have been high enough to elevate serum 25(OH)D concentrations to a level that can decrease cancer risk [37]. In a post-hoc analysis with seven years of follow-up, the intervention versus placebo was not associated with lung cancer incidence (HR=0.86, 95% CI=0.67-1.12) [38]. The potentially suboptimal dose for lung cancer prevention can explain the null association between the trial supplement and lung cancer risk in women with lower baseline serum 25(OH)D concentrations. A large randomized, double-blind, placebo-controlled trial is currently ongoing with a higher dose of vitamin D supplementation (2,000 IU/d) in the primary prevention of cancer and cardiovascular disease [39], and it will need several years to know the effect on lung cancer risk.
Our findings on high versus deficient levels of serum 25(OH)D are in the same direction as a prospective cohort study in the WHI reporting an inverse association of estimated dietary vitamin D exposure with lung cancer risk in never smokers [40]. The majority of lung cancer cases in these two studies were the same group of women. The dietary vitamin D study showed that a higher total vitamin D intake (food+supplements) starting from 600 IU/day was significantly associated with a lower risk of lung cancer among never smokers (HR=0.55 [95% CI=0.31-0.96] for 600 to <800 [41 cases] vs. <100 IU/day; HR=0.37 [95% CI=0.18-0.77] for ≥800 [14 cases] vs. <100 IU/day) [40]. Significant inverse associations were also observed for both NSCLC and adenocarcinoma. In the current study, high 25(OH)D concentrations (≥75 nmol/L) were not associated with lower lung cancer risk (OR=0.76, 95% CI=0.31-1.84, compared with <30 nmol/L). Serum 25(OH)D concentrations at 75 nmol/L or higher have been suggested as the level that may lower cancer risk in general [33]. Given the smaller observed effect size in the serological study than the dietary vitamin D study, we may need a study with a larger number of lung cancer cases to observe a significant association for serum 25(OH)D in relation to lung cancer risk. The biological reasons are unknown for the weaker association of serum 25(OH)D concentrations compared to vitamin D intake with lung cancer risk. In studies examining dietary exposure, it is possible that other dietary factors that correlated with vitamin D intake contributed the lower risk of lung cancer. Also, serum 25(OH)D represents vitamin D exposure from both diet and cutaneous photosynthesis. Thus, serological studies examining serum 25(OH)D are important to confirm the association. Studies considering other components of the vitamin D pathway, such as genetic variants of enzymes, vitamin D protein carrier, and VDR in lung tissue, are needed to elucidate the association.
Although the NSCLC subtype, mainly adenocarcinoma, examined in this study were not significantly associated with serum 25(OH)D concentrations, it is important to consider histological subtypes of lung cancer when we study the risk associated with vitamin D. Vitamin D inhibits lung cancer signaling pathways including epidermal growth factor receptor mutations [41, 42], which are more related to NSCLC than small-cell lung cancer [43]. Also, the expression of VDR is stronger in NSCLC, particularly adenocarcinoma [2], than in small-cell lung cancer [44]. Epidemiological studies suggest that vitamin D exposure is more strongly associated with NSCLC than small-cell lung cancer in smokers [12, 45] and with adenocarcinoma than other subtypes in never smokers [40]. In the U.S. and other parts of the world, the incidence of adenocarcinoma among women is increasing [46], an observation that cannot be entirely attributable to smoking behavior [47]. Research is warranted on low vitamin D intake and serum 25(OH)D concentrations as risk factors for adenocarcinoma.
We observed a trend towards effect modification for serum retinol palmitate/retinol ratio on the association of serum 25(OH)D with NSCLC and adenocarcinoma in the study population. Retinol is crucial to the function of VDR-RXR heterodimer complex by supplying 9-cis-retinoic acid to RXR. However, excess retinol may interrupt the function of VDR-RXR heterodimer by forming RXR-RXR homodimer [6], although it is unclear what levels of vitamin A metabolites would promote the biological change. In the NHANES III, serum retinyl esters concentrations ≥7.0 μg/dL or the retinyl esters/retinol ratio ≥0.08, the level exceeding the capacity of liver storage and potentially causing toxicity [27], were shown to attenuate the inverse association of serum 25(OH)D with lung cancer mortality [10]. In the WHI, a similar pattern was also observed for a high level of dietary and supplemental vitamin A intake [40]. However, in part due to older age, fewer participants in WHI had an excess level of circulating vitamin A than in the NHANES III (15% versus 43% with serum retinyl esters/retinol ratio ≥0.08). Thus, we were unable to use this more biologically meaningful cutoff than the median level to examine the effect modification. The level above which retinol and its metabolites can hinder the function of vitamin D and influence lung cancer risk remains inconclusive and may vary in different populations according to their vitamin A status and other characteristics such as aging.
This study has several strengths. The analysis was conducted on the largest sample, to date, of female never smokers with data on serum 25(OH)D. We used a prospective design so that the bias of reverse causality was unlikely. Study physicians blinded to study components adjudicated the incidence and histology of lung cancer. However, several limitations should be noted. First, we were unable to address potential effects of the long-term systematic or random variation of the biomarkers because they were measured at WHI baseline. In the WHI, serum samples collected five years apart exhibit a modest within-person correlation for 25(OH)D (r=0.61) [48], suggesting that one-point blood measurements moderately represent long-term serum 25(OH)D concentrations. Second, we were unable to adjust for second-hand smoke exposure in all participants because the data were only collected in the OS participants, although the adjustment was unlikely to materially change the study findings, as shown in the sensitivity analysis. Third, in the CaD Trial, the distinction between women assigned to intervention versus placebo may have been reduced, as the trial allowed participants to take personal vitamin D supplement up to 1000 IU/d. Also, in the non-intervention group, some women were ineligible for the CaD Trial because they were taking higher doses, i.e., >600 IU/d, of vitamin D supplements. These potential issues of misclassification may have attenuated our results in the effect modification analysis of the CaD Trial. Lastly, our effect modification analyses should be interpreted as an exploratory work because of the reduced sample size after stratification.
In conclusion, there was no clear association between serum 25(OH)D concentrations and lung cancer risk in a sample of never-smoking postmenopausal women. Future studies should examine effects of higher serum 25(OH)D concentrations (≥75 nmol/L) on adenocarcinoma in never-smoking women.
Supplementary Material
Acknowledgments
We thank Xiaojun Hu and Pamela Y. Yang for their technical assistance with the lab assays and Hui-Chun Tammy Hsu for her statistical assistance. We also thank WHI participants and the Program Office (National Heart, Lung, and Blood Institute, Bethesda, MD) Jacques Rossouw, Shari Ludlam, Joan McGowan, Leslie Ford, and Nancy Geller; (the Clinical Coordinating Center at Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kooperberg; the Investigators and Academic Centers (Brigham and Women’s Hospital, Harvard Medical School, Boston, MA); JoAnn E Manson; (MedStar Health Research Institute/Howard University, Washington, DC) Barbara V Howard; (Stanford Prevention Research Center, Stanford, CA) Marcia L Stefanick; (The Ohio State University, Columbus, OH) Rebecca Jackson; (University of Arizona, Tucson/Phoenix, AZ) Cynthia A Thomson; (University of Florida, Gainesville/Jacksonville, FL) Marian Limacher; (University of Iowa, Iowa City/Davenport, IA) Robert Wallace; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller; (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker.
Financial support: This work was supported by NIH/NCI R03CA167696 and K07CA201334. The WHI program is funded by the National Heart, Lung, and Blood Institute, NIH, US Department of Health and Human Services through contracts HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HSN268201100003C, HHSN268201100004C, and HHSN271201100004C.
Footnotes
Conflict of interest statement: There are no conflicts to disclose.
References
- 1.Samet JM, Avila-Tang E, Boffetta P, et al. Lung cancer in never smokers: clinical epidemiology and environmental risk factors. Clin Cancer Res. 2009;15:5626–45. doi: 10.1158/1078-0432.CCR-09-0376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Menezes RJ, Cheney RT, Husain A, et al. Vitamin D receptor expression in normal, premalignant, and malignant human lung tissue. Cancer Epidemiol Biomarkers Prev. 2008;17:1104–10. doi: 10.1158/1055-9965.EPI-07-2713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hansdottir S, Monick MM, Hinde SL, Lovan N, Look DC, Hunninghake GW. Respiratory epithelial cells convert inactive vitamin D to its active form: potential effects on host defense. Journal of Immunology (Baltimore, Md: 1950) 2008;181:7090–9. doi: 10.4049/jimmunol.181.10.7090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357:266–81. doi: 10.1056/NEJMra070553. [DOI] [PubMed] [Google Scholar]
- 5.Deeb KK, Trump DL, Johnson CS. Vitamin D signalling pathways in cancer: potential for anticancer therapeutics. Nat Rev Cancer. 2007;7:684–700. doi: 10.1038/nrc2196. [DOI] [PubMed] [Google Scholar]
- 6.Carlberg C, Bendik I, Wyss A, et al. Two nuclear signaling pathways for vitamin D. Nature. 1993;361:657–60. doi: 10.1038/361657a0. [DOI] [PubMed] [Google Scholar]
- 7.Pereira F, Larriba MJ, Munoz A. Vitamin D and colon cancer. Endocr Relat Cancer. 2012;19:R51–71. doi: 10.1530/ERC-11-0388. [DOI] [PubMed] [Google Scholar]
- 8.Kilkkinen A, Knekt P, Heliövaara M, et al. Vitamin D status and the risk of lung cancer: a cohort study in Finland. Cancer Epidemiol Biomarkers Prev. 2008;17:3274–8. doi: 10.1158/1055-9965.EPI-08-0199. [DOI] [PubMed] [Google Scholar]
- 9.Freedman DM, Looker AC, Abnet CC, Linet MS, Graubard BI. Serum Vitamin D and Cancer Mortality in the NHANES III Study (1988-2006) Cancer Res. 2010;70:8587–97. doi: 10.1158/0008-5472.CAN-10-1420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cheng TY, Neuhouser ML. Serum 25-hydroxyvitamin D, interaction with vitamin A and lung cancer mortality in the U.S. population. Cancer Causes Control. 2012;23:1557–65. doi: 10.1007/s10552-012-0033-8. [DOI] [PubMed] [Google Scholar]
- 11.Giovannucci E, Liu Y, Rimm EB, et al. Prospective study of predictors of vitamin D status and cancer incidence and mortality in men. Journal of the National Cancer Institute. 2006;98:451–9. doi: 10.1093/jnci/djj101. [DOI] [PubMed] [Google Scholar]
- 12.Weinstein SJ, Yu K, Horst RL, Parisi D, Virtamo J, Albanes D. Serum 25-hydroxyvitamin d and risk of lung cancer in male smokers: a nested case-control study. PLoS One. 2011;6:e20796. doi: 10.1371/journal.pone.0020796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Skaaby T, Husemoen LL, Thuesen BH, et al. Prospective population-based study of the association between serum 25-hydroxyvitamin-D levels and the incidence of specific types of cancer. Cancer Epidemiol Biomarkers Prev. 2014;23:1220–9. doi: 10.1158/1055-9965.EPI-14-0007. [DOI] [PubMed] [Google Scholar]
- 14.Ordonez-Mena JM, Schottker B, Haug U, et al. Serum 25-hydroxyvitamin d and cancer risk in older adults: results from a large German prospective cohort study. Cancer Epidemiol Biomarkers Prev. 2013;22:905–16. doi: 10.1158/1055-9965.EPI-12-1332. [DOI] [PubMed] [Google Scholar]
- 15.Ordonez-Mena JM, Schottker B, Fedirko V, et al. Pre-diagnostic vitamin D concentrations and cancer risks in older individuals: an analysis of cohorts participating in the CHANCES consortium. Eur J Epidemiol. 2015 doi: 10.1007/s10654-015-0040-7. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- 16.Afzal S, Bojesen SE, Nordestgaard BG. Low Plasma 25-Hydroxyvitamin D and Risk of Tobacco-Related Cancer. Clin Chem. 2013;59:771–80. doi: 10.1373/clinchem.2012.201939. [DOI] [PubMed] [Google Scholar]
- 17.Wong YY, Hyde Z, McCaul KA, et al. In older men, lower plasma 25-hydroxyvitamin D is associated with reduced incidence of prostate, but not colorectal or lung cancer. PLoS One. 2014;9:e99954. doi: 10.1371/journal.pone.0099954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chen GC, Zhang ZL, Wan Z, et al. Circulating 25-hydroxyvitamin D and risk of lung cancer: a dose-response meta-analysis. Cancer Causes Control. 2015;26:1719–28. doi: 10.1007/s10552-015-0665-6. [DOI] [PubMed] [Google Scholar]
- 19.The Women’s Health Initiative Study Group. Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials. 1998;19:61–109. doi: 10.1016/s0197-2456(97)00078-0. [DOI] [PubMed] [Google Scholar]
- 20.Jackson RD, LaCroix AZ, Gass M, et al. Calcium plus vitamin D supplementation and the risk of fractures. N Engl J Med. 2006;354:669–83. doi: 10.1056/NEJMoa055218. [DOI] [PubMed] [Google Scholar]
- 21.Wactawski-Wende J, Kotchen JM, Anderson GL, et al. Calcium plus vitamin D supplementation and the risk of colorectal cancer. N Engl J Med. 2006;354:684–96. doi: 10.1056/NEJMoa055222. [DOI] [PubMed] [Google Scholar]
- 22.Hays J, Hunt JR, Hubbell FA, et al. The Women’s Health Initiative recruitment methods and results. Ann Epidemiol. 2003;13:S18–77. doi: 10.1016/s1047-2797(03)00042-5. [DOI] [PubMed] [Google Scholar]
- 23.Curb JD, McTiernan A, Heckbert SR, et al. Outcomes ascertainment and adjudication methods in the Women’s Health Initiative. Ann Epidemiol. 2003;13:S122–8. doi: 10.1016/s1047-2797(03)00048-6. [DOI] [PubMed] [Google Scholar]
- 24.Luo J, Margolis KL, Wactawski-Wende J, et al. Association of active and passive smoking with risk of breast cancer among postmenopausal women: a prospective cohort study. BMJ. 2011;342:d1016. doi: 10.1136/bmj.d1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bondy SJ, Victor JC, Diemert LM. Origin and use of the 100 cigarette criterion in tobacco surveys. Tob Control. 2009;18:317–23. doi: 10.1136/tc.2008.027276. [DOI] [PubMed] [Google Scholar]
- 26.Krasinski SD, Russell RM, Otradovec CL, et al. Relationship of vitamin A and vitamin E intake to fasting plasma retinol, retinol-binding protein, retinyl esters, carotene, alpha-tocopherol, and cholesterol among elderly people and young adults: increased plasma retinyl esters among vitamin A-supplement users. Am J Clin Nutr. 1989;49:112–20. doi: 10.1093/ajcn/49.1.112. [DOI] [PubMed] [Google Scholar]
- 27.Ballew C, Galuska D, Gillespie C. High serum retinyl esters are not associated with reduced bone mineral density in the Third National Health And Nutrition Examination Survey, 1988-1994. J Bone Miner Res. 2001;16:2306–12. doi: 10.1359/jbmr.2001.16.12.2306. [DOI] [PubMed] [Google Scholar]
- 28.Bauer SR, Richman EL, Sosa E, et al. Antioxidant and vitamin E transport genes and risk of high-grade prostate cancer and prostate cancer recurrence. Prostate. 2013;73:1786–95. doi: 10.1002/pros.22717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP, Agurs-Collins T. Measurement characteristics of the Women’s Health Initiative food frequency questionnaire. Annals of Epidemiology. 1999;9:178–87. doi: 10.1016/s1047-2797(98)00055-6. [DOI] [PubMed] [Google Scholar]
- 30.Patterson RE, Levy L, Tinker LF, Kristal AR. Evaluation of a simplified vitamin supplement inventory developed for the Women’s Health Initiative. Public Health Nutr. 1999;2:273–6. doi: 10.1017/s1368980099000361. [DOI] [PubMed] [Google Scholar]
- 31.Wang Y, Jacobs EJ, McCullough ML, et al. Comparing methods for accounting for seasonal variability in a biomarker when only a single sample is available: insights from simulations based on serum 25-hydroxyvitamin d. Am J Epidemiol. 2009;170:88–94. doi: 10.1093/aje/kwp086. [DOI] [PubMed] [Google Scholar]
- 32.Institute of Medicine. Dietary Reference Intake for Calcium and Vitamin D. Washington, DC: The National Academics Press; 2011. [Google Scholar]
- 33.Bischoff-Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson-Hughes B. Estimation of optimal serum concentrations of 25-hydroxyvitamin D for multiple health outcomes. Am J Clin Nutr. 2006;84:18–28. doi: 10.1093/ajcn/84.1.18. [DOI] [PubMed] [Google Scholar]
- 34.Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol. 1993;138:923–36. doi: 10.1093/oxfordjournals.aje.a116813. [DOI] [PubMed] [Google Scholar]
- 35.Cheng TY, Millen AE, Wactawski-Wende J, et al. Vitamin D intake determines vitamin d status of postmenopausal women, particularly those with limited sun exposure. J Nutr. 2014;144:681–9. doi: 10.3945/jn.113.183541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Schnatz PF, Jiang X, Vila-Wright S, et al. Calcium/vitamin D supplementation, serum 25-hydroxyvitamin D concentrations, and cholesterol profiles in the Women’s Health Initiative calcium/vitamin D randomized trial. Menopause. 2014;21:823–33. doi: 10.1097/GME.0000000000000188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Giovannucci E. Can vitamin D reduce total mortality? Arch Intern Med. 2007;167:1709–10. doi: 10.1001/archinte.167.16.1709. [DOI] [PubMed] [Google Scholar]
- 38.Brunner RL, Wactawski-Wende J, Caan BJ, et al. The Effect of Calcium plus Vitamin D on Risk for Invasive Cancer: Results of the Women’s Health Initiative (WHI) Calcium Plus Vitamin D Randomized Clinical Trial. Nutr Cancer. 2011;63:827–41. doi: 10.1080/01635581.2011.594208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Pradhan AD, Manson JE. Update on the vitamin D and omega-3 trial (VITAL) J Steroid Biochem Mol Biol. 2016;155:252–6. doi: 10.1016/j.jsbmb.2015.04.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Cheng TY, Lacroix AZ, Beresford SA, et al. Vitamin D intake and lung cancer risk in the Women’s Health Initiative. Am J Clin Nutr. 2013;98:1002–11. doi: 10.3945/ajcn.112.055905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zhang Q, Kanterewicz B, Shoemaker S, et al. Differential response to 1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)) in non-small cell lung cancer cells with distinct oncogene mutations. J Steroid Biochem Mol Biol. 2013;136:264–70. doi: 10.1016/j.jsbmb.2012.09.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Tong WM, Hofer H, Ellinger A, Peterlik M, Cross HS. Mechanism of antimitogenic action of vitamin D in human colon carcinoma cells: relevance for suppression of epidermal growth factor-stimulated cell growth. Oncol Res. 1999;11:77–84. [PubMed] [Google Scholar]
- 43.Herbst RS, Heymach JV, Lippman SM. Lung cancer. N Engl J Med. 2008;359:1367–80. doi: 10.1056/NEJMra0802714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Kaiser U, Schilli M, Wegmann B, et al. Expression of vitamin D receptor in lung cancer. J Cancer Res Clin Oncol. 1996;122:356–9. doi: 10.1007/BF01220803. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Cheng TY, Goodman GE, Thornquist MD, et al. Estimated intake of vitamin D and its interaction with vitamin A on lung cancer risk among smokers. Int J Cancer. 2014;135:2135–45. doi: 10.1002/ijc.28846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Cheng TY, Cramb SM, Baade PD, Youlden DR, Nwogu C, Reid ME. The International Epidemiology of Lung Cancer: Latest Trends, Disparities, and Tumor Characteristics. J Thorac Oncol. 2016;11:1653–71. doi: 10.1016/j.jtho.2016.05.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Sun S, Schiller JH, Gazdar AF. Lung cancer in never smokers–a different disease. Nat Rev Cancer. 2007;7:778–90. doi: 10.1038/nrc2190. [DOI] [PubMed] [Google Scholar]
- 48.Meng JE, Hovey KM, Wactawski-Wende J, et al. Intraindividual variation in plasma 25-hydroxyvitamin D measures 5 years apart among postmenopausal women. Cancer Epidemiol Biomarkers Prev. 2012;21:916–24. doi: 10.1158/1055-9965.EPI-12-0026. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.